Architecture tests

[Gianeh]

Hi everyone, I really appreciate the work done both by the proposed paper and within code.
I'm trying to reproduce the performance showed in the paper for a Thesis work at University.

I managed to run the code, though I must consider I needed some slight modifications to crucial parts, probably due to compatibility with my machines, but now I'm trying to train different weights for the model using previous architechtures like LapSRN, Bicubic and so on. I encounter an error with shapes of used tensors, both using --mask parameter and without it.

The exception is as follows (removing --mask):

return torch.max_pool2d(input, kernel_size, stride, padding, dilation, ceil_mode)
RuntimeError: Given input size: (256x1x4). Calculated output size: (256x0x2). Output size is too small

I wonder if I'm missing something in my setup or if those archs are intended for testing only... Thank you for the help!